Acta Herpetologica - 2006 - 1 - Morphometric differences between extant and extinct Italian populations of the adder, «Vipera berus» (Linnaeus, 1758) Morphometric differences between extant and extinct Italian populations of the adder, Vipera berus (Linnaeus, 1758) Augusto Gentilli 1, Stefano Scali 2, and Roberto Sacchi 1 1 Laboratorio di Eco-Etologia, Dipartimento di Biologia Animale, Università di Pavia, P.zza Botta 9, 27100 Pavia, Italy 2 Museo Civico di Storia Naturale C.so Venezia 55, 20121 Milano, Italy Vipera berus (Linnaeus, 1758) is the terrestrial snake showing the widest distribu- tion in the world, occuring from Great Britain, France and Northern Italy to the Sakhalin Island and North Korea (Nilson, 1980; Saint Girons, 1980; Nilson et al., 1994; Nilson & Andrén, 1997a). However, adders do not occur uniformly over their distribution area, but are scattered in several isolated populations (Nilson & Andrén, 1997a). Frequently, ecolog- ical traits of borderline and isolated populations differ from those living in the core area of the distribution range of the species, and might be subjected to higher risks of stochas- tic extinction and higher differentiation rates (Mayr, 1970). For example, meadow vipers (Vipera ursinii) show a highly fragmented distribution, many of isolated groups being dif- ferent subspecies (Nilson & Andrén, 1997b, 2001). In Italy Vipera berus inhabits the eastern and central Alps, occurring in most of the Po plain until the nineteenth century (Fig. 1, Bruno, 1992; Societas Herpetologica Itali- ca, 1996). The extinction of the Po plain populations might be caused by land reclama- tion occurred between the end of 18th century and 1930-1940. The progressive heating occurred over the same period (Ghezzi & Riva, 1989; Camuffo, 1990) might have contrib- uted to the extinction of the Po plain adders, favoring more termophilic species, such as V. aspis (Saint Girons, 1975). Po plain adders inhabited only wetlands and marshlands (Ninni, 1879; Camerano, 1888), while alpine adders regularly frequent dry and rocky habitats, generally above 1000 asl (Nilson & Andrén, 1997a; Lapini et al., 1999; Caldonazzi et al., 2002). Moreover, any historical or contemporary record of this species in the pre-alpine area is known (Fig. 1) (Massalongo, 1854, 1859; de Betta, 1857, 1874, 1880; Lessona, 1877, 1879; Bruno, 1992; Societas Herpetologica Italica, 1996), probably because both climatic regime (Tommasel- li et al., 1973) and habitat features of this area do not favour adders. Therefore, Po plain populations might have been geographically isolated from the alpine ones. The isolation of the Po plain population from the main distribution range of the species leads to hypothize that Po plain adders might belong to a different form or subspecies. In this work we investigated this hypothesis, looking for differences in morphometric features between Po plain and alpine adders. Acta Herpetologica 1: 65-71, 2006 66 A. Gentilli et alii We collected data for 28 extant and 28 extinct adders from Italian natural history museums (n =50, 89.2% of specimens, see Appendix 1 for details), and literature (n =6, 10.8% of specimens, Camerano 1888; Boulenger 1896). All the extinct adders from the Po plain preserved in Italian museums were considered. Since adders are sexually dimorphic (Saint Girons, 1978; Scali & Gentilli, 1998), males and females were analysed separately: the Po plain group included 15 males and 13 females, while the alpine one included 12 males and 16 females. In order to describe head morphology, for each specimen we considered (see Scali & Gentilli, 1998 for details): i) the number of postocular scales, ii) the number of front dor- sal head scales, iii) the number of subocular scale rows, iv) head length and width (mm), v) the length and width of the right parietal scale (mm), and vi) length and width of the frontal scale (mm). Since head width correlated with head length (males: rs = 0.45, n = 22, P = 0.034; females: rs = 0.52, n = 27, P = 0.006), we replaced the head width by the correspondent unstandardized residuals versus head length. Since the length and width of frontal and parietal scales were intercorrelated (rs ranging from 0.48 to 0.88 for males and from 0.42 to 0.69 for females), we used a principal component analysis (PCA) to reduce them to a “head scale size” component. For both sexes, PCA extracted only one compo- nent (eigenvalue >1), which accounted for 77.2% and 59.5% of the cumulative variance for males and females respectively, and was positively related to each one of the head scale measures (all variables entered PCA with component loadings higher than 0.75 for males Fig. 1. distribution of extant (l) and extinct (U) Vipera berus in Italy; (Bruno, 1992; Societas Herpetolo- gica Italica 1996, modified by the authors). 67Extant and extint Vipera berus and 0.63 for females). The scores of this “head scale size” component for each adder were used to examine if the size of head scales differed between extinct and extant adders. In order to describe body morphology, we considered body length (from the head tip to the vent, mm) and the number of ventral scales. Since these variables were highly cor- related both in males (rs = 0.60, n = 21, P = 0.004) and females (rs = 0.37, n = 27, P = 0.05), we used a principal component analysis to reduce them to a single “body morphol- ogy” component. For both males and females, PCA extracted only one component (eigen- value >1), which accounted for 69.2% and 71.5% of the cumulative variance for males and females respectively, and was positively related both to the body length and to the number of ventral scales (the two variables entered PCA with component loading higher than 0.83 for both sexes). The scores of this “body morphology” component for each adder were used to examine if the body size differed between extinct and extant adders. Finally, tail features were measured for each specimen using the unstandardized resid- uals of the tail length versus body length, and the number of subcaudal scales. Since these variables were highly correlated (males: rs = 0.46, n = 20, P = 0.039; females: rs = 0.82, n = 27, P<0.001), we used a principal component analysis to reduce them to a single “tail morphology” component. For both males and females, PCA extracted only one compo- nent (eigenvalue >1), which accounted for 71.3% and 94.2% of the cumulative variance for males and females respectively, and was positively related either to the residuals of tail length as to the number of subcaudal scales (the two variables entered PCA with com- ponent loadings higher than 0.84 for males and 0.97 for females). The scores of this “tail morphology” component for each adder were used to investigate if tail features differed between extinct and extant adders. We could not collect all measures from each specimens because their non optimal conservation, so sample sizes varied in different analyses (Tab. 1). Correlation Analysis (Spearman’s rho coefficient) was used to investigate relation- ships among variables. We used the Bonferroni correction for multiple tests to adjust the observed significance level to the number of comparisons (Rice, 1989): in a set of K simul- taneous comparisons, a given test was considered to be statistically significant at 0.05 level only if it was significant also at the 0.05/K level. We used discriminant Function Analysis (dFA, stepwise forward method) using all morphometric measures as independent vari- ables and collecting area (Alps vs Po plain) as classifying factor, in order to determine i) if Po plain (extinct) and alpine (extant) adders differed on the basis of their morphological features, and ii) what morphological features, if any, accounted for such a difference. All analyses were performed using SPSS rel. 6.1.2 software package (1996). Head, body, and tail variables were not intercorrelated after Bonferroni correction (K = 28; P < 0.00178 for relationships to be significant). discriminant function analysis for males generated a high significant discriminant function (19.9, P < 0.0001, λ = 0.26), which correctly classified 90% of specimens. The stepwise procedure extracted only two variables: the number of postocular scales and the tail morphology. The males of Po plain adders had more postocular and subcaudal scales and longer tails than alpine ones (Tab. 1). Similarly, dFA for females generated a highly significant discriminant function (27.6, P < 0.0001, λ = 0.15), which correctly classified 95.6% of specimens. The stepwise proce- dure extracted three variables: the number of postocular scales, the tail morphology, and the 68 A. Gentilli et alii Ta bl e 1. M or ph om et ri c m ea su re s of P o pl ai n an d al pi ne a dd er s. M al es Fe m al es Po p la in A lp s Po p la in A lp s n M in M ax M ea n Sd n M in M ax M ea n Sd n M in M ax M ea n Sd n M in M ax M ea n Sd H ea d m or ph ol og y Po st o cu la r sc al es ( n) 11 3 5 3. 7 0. 8 12 2 4 2. 8 0. 6 9 3 5 3. 9 0. 8 15 2 4 3. 2 0. 6 Fr on t do rs al h ea d sc al es ( n) 8 7 13 10 .5 2. 2 12 6 22 11 .6 5. 2 11 11 30 17 .6 5. 6 15 5 27 14 .5 5. 0 Su bo cu la r ro w s (n ) 11 1 2 1. 2 0. 4 12 1 2 1. 1 0. 3 13 1 2 1. 6 0. 5 16 1 2 1. 2 0. 4 H ea d w id th ( m m ) 10 10 .7 15 .9 13 .6 1. 74 12 9. 2 15 .3 12 .8 1. 78 13 9. 2 19 .0 14 .0 2. 9 14 11 .8 18 .2 14 .9 1. 91 H ea d le ng th ( m m ) 10 19 .3 25 .8 22 .5 1. 91 12 16 .2 29 .2 21 .5 3. 58 13 15 .3 34 .6 24 .1 4. 8 14 18 .1 28 .0 23 .2 2. 44 Pa ri et al s ca le le ng th ( m m ) 10 3. 5 6. 2 4. 8 0. 84 12 3. 0 4. 8 3. 9 0. 62 10 2. 6 5. 2 3. 9 0. 8 13 2. 7 5. 5 4. 4 0. 84 Pa ri et al s ca le w id th ( m m ) 10 2. 2 3. 3 2. 6 0. 38 12 1. 6 2. 9 2. 2 0. 40 10 1. 6 3. 0 2. 2 0. 4 13 1. 7 3. 3 2. 5 0. 47 Fr on ta l s ca le le ng th ( m m ) 10 3. 7 5. 7 4. 6 0. 70 12 2. 4 4. 6 3. 6 0. 71 8 2. 7 4. 5 3. 7 0. 5 15 2. 7 5. 7 4. 6 0. 79 Fr on ta l s ca le w id th ( m m ) 10 2. 6 3. 7 3. 3 0. 36 12 1. 8 4. 6 3. 0 0. 78 8 2. 3 3. 7 2. 9 0. 4 15 2. 7 3. 6 3. 1 0. 29 B od y m or ph ol og y B od y le ng th ( m m ) 10 31 1. 4 53 6. 0 43 7. 0 67 .2 7 12 16 6. 8 41 2. 6 32 6. 1 85 .2 3 13 17 3. 7 69 0. 0 42 0. 9 13 3. 8 15 30 6. 9 53 2. 0 43 6. 1 56 .3 3 V en tr al s ca le s (n ) 15 13 6 14 8 14 2. 1 3. 72 12 13 6 14 9 14 1. 0 3. 63 13 14 0 15 4 14 5. 9 4. 6 16 13 8 15 5 14 6. 4 5. 20 Ta il m or ph ol og y Ta il le ng th ( m m ) 10 44 .0 92 .0 76 .3 15 .7 8 11 30 .0 74 .0 56 .0 15 .3 1 13 28 .0 10 0. 0 56 .6 20 .4 16 34 .0 58 .0 49 .3 7. 07 Su bc au da l s ca le s (n ) 14 35 47 41 .5 3. 50 11 33 41 36 .8 2. 66 12 29 42 33 .8 3. 7 16 18 31 26 .8 3. 49 69Extant and extint Vipera berus number of subocular scale rows. The females of Po plain adders had more postocular scales, more subocular scale rows, more subcaudal scales, and tails longer than alpine ones (Tab. 1). In conclusion, our results showed morphological differences between Po plain and alpine adders, supporting our hypothesis that Po plain and alpine populations of this spe- cies might be considered two different ecotypes or subspecies. However, our morphologi- cal analyses have to be considered only a preliminary step, and genetic analyses are need- ed to better clarify the sistematic position of Po plain adders. Further analyses would also check for relationships between Po plain adders and V. b. bosniensis, since their distribu- tion range might overlap in the coastal wetlands of North Adriatic. ACKNoWLEdGEMENTS We thank the directors and curators of the museums for allowing us to measure the speci- mens analysed in the present paper. We also thank dr. Edoardo Razzetti for his useful comments to the manuscript. REFERENCES Boulenger, G.A. (1896): Catalogue of the snakes in the British Museum (Natural History). Taylor and Francis Ltd, London. Bruno, S. (1992): Repertorio zoogeografico, geonemico, tassonomico, biografico e biblio- grafico degli studiosi e degli studi di erpetologia italiana. I. Serpentes: 1800-1899. Atti Acc. Rov. Agiati a. 241 (1991), Serie VII 1, B: 5-256. Caldonazzi, M., Pedrini, P., Zanghellini, S. (2002): Atlante degli anfibi e dei rettili della provincia di Trento (Amphibia, Reptilia), 1987-1996 con aggiornamenti al 2001. St. trent. Sci. nat. Acta Biol. 77: 1-173. Camerano, L. (1888): Monografia degli ofidi italiani. Parte prima: Viperidi. Mem. R. Accad. Sci. Torino, Serie II 39: 195-243. Camuffo, d. (1990): Clima e uomo. Garzanti, Milano. de Betta, E. (1857): Erpetologia delle province venete e del Tirolo meridionale. Atti Mem. Acc. Agr. Sci. Lett., Verona 35: 1-365. de Betta, E. (1874): Rettili ed Anfibi. Parte IV. In: Fauna d’Italia divisa in due branche Vertebrati ed Invertebrati, p. 1-107. Vallardi, Milano. de Betta, E. (1880): Sulla distribuzione geografica dei serpenti velenosi in Europa e parti- colarmente nell’Italia. Atti R. Ist. ven. Sci. Lett. Arti, Venezia 6: 359-392. Ghezzi, A., Riva, I. (1989): Il clima del territorio delle province di Cremona e Mantova. Pianura 3: 29-46. Lapini, L., dell’Asta, A., Bressi, N., dolce, S., Pellarini, P. (1999): Atlante corologico degli anfibi e dei rettili del Friuli-Venezia Giulia. Museo Friulano di Storia Naturale, Udine. Lessona, M. (1877): delle vipere in Piemonte. Atti R. Accad. Sci. Torino (Reprint) 12: 1- 10. 70 A. Gentilli et alii Lessona, M. (1879): Nota intorno al Pelias berus in Piemonte. Atti R. Accad. Sci. Torino (Reprint) 14: 1-2. Massalongo, A. (1854): Saggio di un’erpetologa popolare veronese. Tip. Giuseppe Antonel- li, Verona. Massalongo, A. (1859): Catalogo dei rettili delle province venete. Atti R. Ist. Ven. Sci. Lett. Arti, Venezia 4: 300-310. Mayr, E. (1970): L’evoluzione delle specie animali. Einaudi Ed., Torino. Nilson, G. (1980): Male reproductive cycle of the European adder, Vipera berus and its relation to annual activity periods. Copeia 4: 729-737. Nilson, G., Andrén, C., Szyndlar, Z. (1994): The systematic position of the common adder, Vipera berus (L.) (Reptilia, Viperidae), in North Korea and adjacent regions. Bonn. Zool. Beitr. 45: 49-56. Nilson, G., Andrén, C. (1997a): Vipera berus (Linnaeus, 1758). In: Atlas of Amphibians and Reptiles in Europe, P. 388-389. Gasc, J.P., Cabela, A., Crnobrnja-Isailovic, J., Haffner, P., Lescure, J., Martens, H., Martínez Rica, J.P., Maurin, H., oliveira, T., Sofianidou, T.S., Veith, M., Zuiderwijk, A., Eds., Societas Europea Herpetologica, Muséum National d’Histoire Naturelle (IEGP/SPN), Paris. Nilson, G., Andrén, C. (1997b): Vipera ursinii (Bonaparte, 1835). In: Atlas of Amphibi- ans and Reptiles in Europe, p. 400-401. Gasc, J.P., Cabela, A., Crnobrnja-Isailovic, J., Haffner, P., Lescure, J., Martens, H., Martínez Rica, J.P., Maurin, H., oliveira, T., Sofianidou, T.S., Veith, M., Zuiderwijk, A., Eds., Societas Europea Herpetologica, Muséum National d’Histoire Naturelle (IEGP/SPN), Paris. Nilson, G., Andrén, C. (2001): The meadow and steppe vipers of Europe and Asia – the Vipera (Acridophaga) ursinii complex. Acta zool. hung. 47: 87-267. Ninni, A.P. (1879): Breve nota intorno al marasso Vipera (Pelias) berus L. nel Veneto. Atti Soc. Ital. Sci. Nat., Milano 22: 175-181. Rice, W.R. (1989): Analyzing tables of statistical tests. Evolution 43: 223-225. Saint Girons, H. (1975) : Coexistence de Vipera aspis et de Vipera berus en Loire-Atlan- tique: un probleme de competition interspecifique. La Terre et la Vie 29: 590-613. Saint Girons, H. (1978): Morphologie externe comparée et systématique des Vipères d’Europe (Reptilia, Viperidae). Revue Suisse Zool. 85: 565-595. Saint Girons, H. (1980): Biogéographie et évolution de vipères européennes. C.R. Soc. Biogéogr. 496: 146-172. Scali, S., Gentilli, A. (1998): Morphometric analysis, sexual dimorphism and distribution of extinct adders (Vipera berus) of the Po Plane (Northern Italy). In: Current studies in Herpetology, p. 391-396. Miaud, C., Guyétant, R., Eds. Le Bourget du Lac, France, S.E.H. Societas Herpetologica Italica (1996): Atlante provvisorio degli Anfibi e dei Rettili italiani. Ann. Mus. Civ. St. nat. «G. doria», Genova 91: 95-178. SPSS (1996): Statistical Package for Social Science for Windows rel. 6.1. SPSS Inc., Chica- go. Tommaselli, R., Balduzzi, A., Filipello, S. (1973): Carta bioclimatica d’Italia. Collana Verde 33, Ministero dell’Agricoltura, Roma. 71Extant and extint Vipera berus Appendix 1. List of specimens. Museum Collection code Museo Civico di Storia Naturale (Milano) 1052; 1529; 1530; 1531; 1532; 1533; 1534; 1535; 1536; 1578A; 1578B; 1591; 1592; 1623; 3286; 3579; 3749; 3758; 3760 Museo Civico di Storia Naturale (Morbegno) 3; 4; 26; 29; 41; 65; 67; 1f Unknown number Museo Regionale di Scienze Naturali (Torino) R283; R522 Museo Civico di Storia Naturale (Verona) 866; 867; 868 (2 specimens); 869 (2 specimens); 870; 872; 873a; 873b; 873c; 873d; 887 (2 specimens); 1330 Museo Zoologico “G. Scarpa” (Treviso) 2mm Unknown numbers Museo Zoologico dell’Università (Firenze) 12062; 12101-02 Museo Zoologico dell’Università (Pavia) 230/476; 230/477